ykai-23 - Tumblr blog

[ykai-23]

Blog 9 Week 11 ‘The algorithmic rise of the “alt-right’

Blog 9 ‘The algorithmic rise of the “alt-right’

While the argument made by Daniels (2008) that algorithms could facilitate the spread of white supremacy ideologies sounds plausible, algorithms do not necessarily have direct causal relations with hate crimes nowadays. The argument that algorithms play a leading role in real-life negative consequences may parallel the filter bubble theorization, arguing that the design of internet platforms, such as search engines or social media, that use algorithm systems to recommend and display personalized information based on users’ preferences only, reinforcing pre-existing beliefs (Pariser, 2011). In the context of social media, Bruns (2021) argues that individuals with different political views could still come across their opponents’ tweets deliberately or serendipitously through hashtag affordance on Twitter. He further reasons that “social media users would also continue to use the same platforms for a variety of other purposes”, leading them to encounter wide-ranging users with different views instead of being confined to “one closed space,” (Bruns, 2021, p.40). The tendency of looking at similar viewpoints may be attributed to homophily, which refers to “a tendency for friendships to form between those who are alike in some respect,” (Lazarsfeld, 1954, p.23). Therefore, hate crimes today may be due to long existing racism in the society and human beings’ predisposition to look for like-minded ideas rather than being generated by algorithms solely.

Reference

Bruns, A., (2021). Echo chambers? Filter bubbles? The misleading metaphors that obscure the real problem. In Hate speech and polarization in participatory society (pp. 33-48). Routledge.

Daniels, J., (2018). The algorithmic rise of the “alt-right”. Contexts, 17(1), pp.60-65.

Lazarsfeld, P.F. and Merton, R.K., (1954). Friendship as a social process: A substantive and methodological analysis. Freedom and control in modern society, 18(1), pp.18-66.

Pariser, E., (2011). The filter bubble: How the new personalized web is changing what we read and how we think. Penguin.

[ykai-23]

Blog 8 Week 10 Misinformation and Its Correction

Despite acknowledging a few reasons for how Web 2.0 amplified the circulation of misinformation Lewandowsky et al. (2012) identified several long-existing sources of misinformation including unintended inaccuracies from insufficient information, rumors and fiction, governments, corporate vested interests, and the media prior to the era of the internet. The identification of misinformation sources may debunk the taken-for-granted technological determinist view that the internet is the main culprit for the spreading of misinformation since the dissemination has a long root before the internet era.

The readily circulation of misinformation on the internet might be due to the functionalities of Web 2.0 characterized by participatory creation, sharing, and engagement of User-Generated Content (UGC) (Kaplan and Haenlein, 2010). Thus, internet users could publish content across media platforms without the checks of gatekeepers, namely the journalists or editors, and the veracity and validity of content are subjective to variability (Lewandowsky et al., 2012). Moreover, online information in multimedia formats could be more accessible and straightforward than seeking advice from professionals, since users could search for information within a few clicks, and therefore netizens might seek for advice online despite the information potentially containing misleading messages (Lewandowsky et al., 2012).  These two reasons manifest the role of the internet in exacerbating the proliferation of false information, but the main reason lies in existing sources of misinformation, such as mistakes resulting from deficit in information or different parties vying for vested interests.

Reference

Kaplan, A.M. and Haenlein, M., 2010. Users of the world, unite! The challenges and opportunities of Social Media. Business horizons, 53(1), pp.59-68.

Lewandowsky, S., Ecker, U.K., Seifert, C.M., Schwarz, N. and Cook, J., 2012. Misinformation and its correction: Continued influence and successful debiasing. Psychological science in the public interest, 13(3), pp.106-131.

[ykai-23]

Blog 7 Week 9 A Declaration of the Independence of Cyberspace | Electronic Frontier Foundation

In 1996, John Barlow, a pioneering digital-rights activist, proposed an appealing notion that cyberspace is a free, equal, borderless, and self-governed realm where governments should not have control and power over it, resembling a utopia online (Barlow, 2018). While Barlow claims that liberty online cannot be restricted by governments over the long term worldwide, countries including China, Iran, and Uzbekistan have adopted methods, such as legal frameworks, technological barriers, and censorship, to block the access to certain content online (Zittrain and Palfre,2008). In addition, according to Surfshark (2022), the governments of five countries including Belarus, Iraq, North Korea, Oman, and Tukmenistan have even criminalized VPN usage to further crack down on strategies to bypass internet filtering, restricting citizens’ liberty to access content online.

Apart from free access online, Barlow also asserted that problems online do not exist, and cyberspace is a new civilization being fair and humane. Examples of negative consequences, generated by the spread of disinformation on social media could be in Plateau, Nigeria. In June 2018, a Facebook post including photos of a disfigured baby and a man with the skull open, misleadingly asserted that they were Christians of Beroms killed by Fulanis Muslims and resulting in regional conflicts and the killing of many Fulanis Muslims; in reality, the photos were from an unrelated incident in Congo in 2012 (Apuke & Omar, 2020; Edwin & Yalmi, 2019). Therefore, issues and malevolence could arise from online space and lead to real consequences nowadays. One of the reasons for Barlow’s overtly positive attitudes towards the internet might be due to that limited number of people had access and control to the internet in 1996, whereas today an eclectic range of governments, netizens, and organizations with vested interests online could hardly coexist without experiencing dispute, chaos, and control.

Reference:

Apuke, O. and Omar, B. (2020) ‘Fake news proliferation in Nigeria: Consequences, motivations, and Prevention Through Awareness Strategies’, Humanities & Social Sciences Reviews, 8(2), pp. 318–327. doi:10.18510/hssr.2020.8236.

Barlow, J.P. (2018) A declaration of the independence of Cyberspace, Electronic Frontier Foundation. Available at: https://www.eff.org/cyberspace-independence (Accessed: 12 December 2023). 

Edwin, M.R. and Yalmi, G.L. (2019) ‘Fake news and Home Truths, its effects on the contemporary political narratives: An appraisal of India and Nigeria’, International Journal of Humanities and Social Science, 9(6). doi:10.30845/ijhss.v9n6p4.

SurfShark. (2022). Countries where VPN usage is limited in 2022. Statista. Statista Inc.. Accessed: November 27, 2023. https://www-statista-com.liverpool.idm.oclc.org/statistics/1343693/worldwide-virtual-private-network-usage-limited/

Zittrain, J. and Palfrey, J. (2008) ‘Internet filtering: The politics and mechanisms of Control’, Access Denied, pp. 29–56. doi:10.7551/mitpress/7617.003.0005.

[ykai-23]

Blog 6 Week 8 State Surveillance and Social Democracy: Lessons after the Investigatory Powers Act 2016

The Investigatory Powers Act (IPA) was enacted in the UK to provide a legal basis for data retention, granting the government access to comprehensive communications and the retention of data on internet use and phone calls (Murphy, 2019). The act has been considered as a threat and erosion of the agency of several institutions that sustain the function of democracy. First, the IPA provides officials with warrants to intercept communication between lawyers and their clients in matters relevant to national security or prevention of death. Second, it provides Judicial Commissioners with access to pinpoint journalistic sources without requests for warrants. Last, the act allows the government to tap into trade union membership data and communication between other organizations.

The lack of sufficient confidentiality between lawyers and clients, journalists and sources, and trade unions and subversive organizations generates chilling effects that thwart voices from clients, sources, and workers, creating power imbalance between the government and minorities. Since the function of media and trade unions is fundamental in democracy in relation to informing the public, forming public opinions, and advocating for workers’ rights, the interception granted by IPA could undermine the agency of these institutions to ensure checks and balances, eventually eroding the function of a democracy.

Reference

Murphy, C.C. (2019) ‘State Surveillance and Social Democracy: Lessons after the investigatory powers act 2016’, SSRN Electronic Journal [Preprint]. doi:10.2139/ssrn.3494880.

[ykai-23]

Blog 5 Week 6 'I've Got Nothing to Hide' and Other Misunderstandings of Privacy

The seemingly plausible phrase “I have nothing to hide” mentioned in Solove(2007, p. 745) downplays privacy matters as trivial compared to security issues like terrorism and disregards possible menace behind governments’ unconstrained access to personal data. While the phrase implies that privacy is only a matter of disclosure of personal information, privacy is also closely associated with other concerns, including the courage of speaking up, protection of personal space, and checks and balances of power. 

Ceding control of privacy grants the government significant power to have access to private lives, and unchecked power could lead to corruption. Personal data can be exploited and be leveraged as a bargaining chip for intimidation, influencing personal decision making, and even fabricating disinformation, which might gradually lead to the erosion of free speech and democracy.

In addition, unlimited and unjustified access of personal data lacks contractual spirit including transparency, consent, and equal communication, and deteriorates a rule-based trustworthy society where individual rights are protected. Thus, privacy is closely related to personal security and the aggregate social security, which should not be only controlled by a small group of people. In terms of attacking terrorism, questions that need to further examine are how effective it is to prevent terrorism by granting the government considerable power and what possible alternatives could be explored to strike a balance between preventing terrorism and protecting privacy.

Referece:

Solove, D.J., (2007). I've got nothing to hide and other misunderstandings of privacy. San Diego L. Rev., 44, p.745.

[ykai-23]

Blog 4 Week 5 Explaining Explanations: An Overview of Interpretability of Machine Learning

While methods including linear proxy models, decision trees, automatic-rule extraction, and salience mapping, have been taken to simplify and explain the processing of data in a network, each approach has its limit in terms of lack of faithfulness or scalability to original network due to respective limitations (Gilpin et al. 2018).

Apart from the fragmentation and limitations of each approach, a greater dilemma lies in the tradeoff or a spectrum between interpretability, meaning the understandability of the AI system, and completeness, which is the accurate and detailed description of the system. Understanding the system is imperative to ensure the impartiality of algorithmic system and garner wider recognition for users being able to trust the system they use; however, transparency and accuracy could be undermined when explanations are simplified and this could have life-threatening ramifications for usage in life-critical automated systems, such as autonomous vehicles or medical diagnostic tools.

In order to tackle the constraints of various present approaches and the predicament between intelligibility and entirety, cooperative teamwork among interdisciplinary branches of studies is needed to merge approaches to address the challenge of isolated methodologies in explaining AI systems and thus enhance the system's explainability in considering both dimensions.

Reference

Gilpin, L.H., Bau, D., Yuan, B.Z., Bajwa, A., Specter, M. and Kagal, L., (2018), October. Explaining explanations: An overview of interpretability of machine learning. In 2018 IEEE 5th International Conference on data science and advanced analytics (DSAA) (pp. 80-89). IEEE.

[ykai-23]

Blog 3 Week 4 The Economics of Big Data and Artificial Intelligence

Despite the widespread belief that Big Data and AI could considerably facilitates human work productivity, evidence counterintuitively shows correlations between deceleration in economic growth in the United States and the advancement of big data; however, Big Data solely contribute to the productivity and profitability of companies using analytics (Mihet & Philippon, 2019). These phenomena showcase that Big Data might exclusively benefit tech-savvy companies with large accessible databases, enabling them to make better predictions and generate revenue, rather than benefiting society as a whole.

Frontier companies' exclusive privilege of leveraging large numbers of data to enhance sales may lead to greater stratification in society since they only take, meaning extracting data exhaust and profiting from consumption, from the public without reciprocating benefits to consumers. Furthermore, tech-ready giants might cut down in-house labour and retain only tech-proficient workers, outsourcing manual work to developing countries, which enables them to be more profitable but also potentially exploitative.

Apart from the infocratic and manipulative nature of companies using analytics, they are not omnipotent in controlling consumer behaviour. Users could still navigate their ways to resist discriminatory pricing imposed by Big Data technology by comparing products prices across multiple platforms, seeking input from mutual friends, or simply withdrawing from online shopping completely. 

In a nutshell, Big Data analytics may not contribute to the welfare of human beings in total, and contrastingly, they serve to exacerbate social disparity. However, Big Data do not comprise the whole activities of society, so their power is limited, and consumers still possess limited agency to resist the system.

Reference

Mihet, R. and Philippon, T., (2019). The economics of big data and artificial intelligence. In Disruptive Innovation in Business and Finance in the Digital World (pp. 29-43). Emerald Publishing Limited.

[ykai-23]

Blog 2 Week 3 Big other: surveillance capitalism and the prospects of an information civilization.

The concepts of surveillance capitalism, logic of accumulation, and formal indifference, as propounded by Zuboff (2015), explicate the emerging structure of capitalism formulated by Google, which invades the public’s privacy rights, predicting and modifying user behaviours for generating revenues through surveilling, extracting, and analysing online users’ data exhaust unilaterally.

The seemingly consensual relationship between tech firms and users is essentially a semi-coercive contract since the drafting process of lengthy and esoteric contracts lacks users’ engagement and understanding. Public awareness of this invasion of privacy needs to be raised for generating privacy laws and taking legal actions to counteract the invisible hands of Tech Giants.

Although the infocratic architecture created by tech companies in network society establishes a structure of hegemony requiring attention and actions to mitigate, the notion of Big Other coined by Zuboff (2015) suggesting that media conglomerates wield omnipotent power over the public, resembling a panopticon, might be deemed similar to the concept of media panics that catastrophize the capacity of data analytics in turning the world into a watertight prison.

Users still possess a limited amount of agency and strategies in contemplating and resisting surveillance and indoctrination of expected consumer behaviors. Additionally, even though online activities account for a large portion of our lives, part of our actions and operations remain out of sight of tech companies, which cannot be tracked or monitored, unlike a prisoner in a panopticon.

Reference

Zuboff, S. (2015). Big other: surveillance capitalism and the prospects of an information civilization. Journal of information technology, 30(1), 75-89

[ykai-23]

Blog 1 Week 2 Internet Filtering: The Politics and Mechanisms of Control

The rationale of carrying out internet filtering and content censorship to prevent the circulation of child pornography, the violation of intellectual property, and international terrorism mentioned in the article seems plausible(Zittrain and Palfrey, 2008), but the power of controlling the internet could lead to corruption and violation of human rights easily if not kept in check. For Instance, an Iranian woman, Mahsa Amini, died possibly due to police violence, which sparked protests within Iran. To counterattack the insurgencies of Iranian citizens, the officials in Iran blocked access to remaining available social media and websites to censor videos of police violence and protests from circulation (Hodge, 2022). As shown in the figure, the internet connectivity in Iran was completely shut down on 22 and 23 September, and then it fluctuated between being connected and restricted for almost two weeks (Internet Outage Detection and Analysis, 2022). Apart from this example of filtering online, governments have the ability to synergize with private companies to gain user location data or surveil text messages among users to track, counterattack, and jail dissidents in local gatherings. Although countermeasures could be taken, not all users could be tech-savvy enough to use VPN, end-to-end encryption, or Tor to bypass surveillance and internet filtering, which could take a toll on concerted power to resist injustice, dictatorship, and hegemony. Therefore, citizens should be aware of the power imbalance among government, corporations, and themselves to make informed decisions and ensure that this power meant for protection is not being abused.

(Internet Outage Detection and Analysis, 2022)

Reference

Hodge, R. (2022, October 06). VPN demand surges in Iran as protests continue, study shows. CNET. Retrieved October 10, 2023, from https://www.cnet.com/tech/services-and-software/vpn-demand-surges-in-iran-as-protests-continue-study-shows/

Internet Outage Detection and Analysis. (December 5, 2022). Share of internet connectivity in Iran between September 19th to October 6th, 2022, based on Border Gate Protocol* [Graph]. In Statista. Retrieved October 10, 2023, from https://www.statista.com/statistics/1350215/iran-share-of-internet-connectivity/

Zittrain, J. and Palfrey, J., (2008). Internet filtering: The politics and mechanisms of control. Access denied: The practice and policy of global Internet filtering, 41.